JOURNAL OF CHEST & PULMONARY MEDICINE (ISSN:2732-480X)

Previous Issue  |   Current Issue  |   Archive  |   Recent Articles  |   Publications      
     PDF Download

Current Status of Empyema Management and Outcomes

Brianna L. Siracuse1, Amber L. Turner2, Russell C. Langan3, Andrew B. Nguyen4, Subroto Paul4, Joanna Sesti4* 

1 Rutgers New Jersey Medical School, Newark, NJ 07103, United States
2 Department of Surgery, RWJBarnabas Health, Livingston, NJ 07039, United States
3 Department of Surgical Oncology and Hepatobiliary Surgery, RWJBarnabas Health, Livingston, United States
4 Department of Thoracic Surgery, RWJ Barnabas Health, West Orange, United States

CitationCitation COPIED

Siracuse BL, Turner AL, Langan RC, Nguyen AB, Paul S, et al. Current Status of Empyema Management and Outcomes. J Chest Pulm Med. 2020 May;1(1):105.

Copyright:

2020 Siracuse BL, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 international License, which permits unrestricted use, distribution and reproduction in any medium, provided the original author and source are credited.

Abstract

Objectives
Empyema remains relevant today despite advances medical treatment, and continues to be associated with high degree of morbidity and mortality. Due to the different stages at presentation, a clear treatment algorithm does not exist. Large series have proposed a less invasive treatment involving drainage followed by Alteplase (TPA) and Dornase instillation in lieu of surgical decortication. We sought to analyze modern trends in management and outcomes for empyema.

Methods
Patients with empyema were identified in the State Inpatient Database (SID) of New York, California, Florida, and Washington between January 1, 2006 and December 31, 2014. A comparison of demographic and clinicopathologic features, 30- and 90-day readmissions, and in-hospital mortality was performed. Univariate analyses were performed using chi-square tests, student t tests, and binary logistic regression. Factors determined to be significant in the univariate analysis were included in the multivariate logistic regression analysis model.

Results
69,935 inpatients with empyema were identified. Surgical management of empyema peaked in 2009 (44%) and slowly decreased until 2014 (39%). There was a decrease in overall length of stay, accompanied by a trend towards more discharges to short term rehab or skilled nursing facilities (38 vs 34%, p < 0.001). Mortality rates for empyema were lower in patients receiving surgical intervention or intercostal drainage versus thoracentesis.

Conclusion
In our sample, surgical management of empyema has slowly decreased from 2006 to 2014. This practice change was not associated with changes in hospital mortality but we did note shorter length of stay and overall lower readmissions rates. 

Keywords

Empyema; Pleural effusion; Pneumonia

Abbreviations

TPA: Alteplase
DNAse: Dornase
SID: State Inpatient Database

Introduction

Empyema thoracis is an ancient disease that continues to be a relevant medical issue. The term empyema thoracis comes from the Greek, which means “pus in the chest.” The most common source is bacterial pneumonia which leads to a parapneumonic effusion. Other sources include esophageal rupture, descending mediastinitis, subdiaphragmatic infections with pleural extension, cervical and thoracic spine infections, and post procedural etiologies. Despite the widespread use of broad spectrum antibiotics and chest imaging technologies, empyema remains a common complication of pneumonia and a significant source of morbidity and mortality. There are approximately one million patients hospitalized in the United States each year with pneumonia. Parapneumonic effusions will develop in 20% to 40% of these patients. Of those patients who develop a parapneumonic effusion, 5% to 10% will develop an empyema [1].

Both surgical and non-surgical treatment strategies exist for empyema. Observation alone is recommended for small, free-flowing parapneumonic effusions when decubitus chest radiographs demonstrate free-flowing collections < 1 cm in width [1]. Thoracentesis, tube thoracostomy with and without fibrinolytics, and surgical decortication are the most commonly used treatment methods. Several guidelines have been published to aid providers in the treatment algorithm for empyema including The British Thoracic Society, American College of Chest Physicians, and the American Association of Thoracic Surgeons [1-3]. Overall, there has been a trend towards less invasive management with emphasis on thoracostomy tube drainage and increasing use of Alteplase (TPA) and Dornase (DNAse). The aim of this analysis was to look at current trends in the management of empyema and associated outcomes. 

Methods

Study Design
Patients with empyema were identified using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD9-CM) diagnosis codes in the State Inpatient Database (SID) of New York, California, Florida, and Washington, a part of the Healthcare Cost and Utilization Project, from the Agency for Healthcare Research and Quality. The SID is an all-payer administrative database that collects hospital inpatient discharge information. In the database, each patient is given a unique numerical identifier that enables tracking of patient admissions within that state for a given year. These states were chosen to provide regional variety and a large sample size for this study. Patients who were diagnosed with empyema were identified using the diagnostic ICD-9-CM code 510.0, 510.9, or 511.1 between January 1, 2006 and December 31, 2014. Patients under 18 and those without a variable to track readmissions were excluded from this study (Figure 1). This study was approved by our institutional review board, IRB protocol No. 17-67 on November 14, 2017 (subsequently renewed on November 13, 2018).

Definitions
Diagnostic codes for causal microorganisms within the limitations of the ICD-9-CM were identified, as follows: staphylococcus species (including methicillin-resistant species), Streptococcus species, gram-negative bacilli, fungi, and unknown (which included culturenegative and non-coded cases). Treatment modalities were separated into surgery, intercostal drainage, and thoracentesis. Patients who underwent surgery were identified using the following procedural ICD-9 CM codes: 340.9, 345.1, and 345.2. ICD-9-CM codes for intercostal drainage were 340.4 and 340.6. Intercostal drainage was identified by the ICD-9-CM code 349.1. Specific ICD-9-CM codes and definitions are found in Appendix Table 1.


Figure 1: Flowchart of Data Exclusions Used to Examine Empyema Trends


Appendix Table 1: ICD-9 Codes used for Analysis of Empyema Trends

Statistical Analysis

Demographic and clinicopathologic factors were abstracted for each patient. Demographic information included age, gender, race, median income quartile, primary insurance payer, and length of stay. Clinicopathologic factors were anemia, congestive heart failure, chronic pulmonary disease, diabetes, hypertension, liver disease, metastatic cancer, obesity, peripheral vascular disorder, pulmonary circulation disorder, renal failure, solid tumor without metastasis, primary microorganism, and diagnosis upon readmission. Charlson and Elixhauser comorbidity scores were also calculated for each patient [4,5].

A comparison of demographic and clinicopathologic features was performed between 2006 and 2014 to determine if there were any changes in the population of patients who were diagnosed with empyema over time. 30- and 90-day readmission rates were defined as any subsequent admission within 30 or 90 days after the initial discharge date, respectively. Mortality rate during initial hospitalization were also calculated. Additionally, yearly trends in readmission and mortality were analyzed by Charlson and Elixhauser comorbidity score. Univariate analyses were performed using chi-square tests, student t tests, and binary logistic regression. A p-value of <0.01 was considered significant. Factors determined to be significant in the univariate analysis were included in the multivariate logistic regression analysis model. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated to determine the strength of association within each subgroup. Mortality rates by primary microorganism and treatment modality by year were also calculated. All statistical analyses were performed using SAS Statistical Software, version 9.3 (SAS Institute Inc.; Cary, NC). As per SID reporting guidelines, values of <10 were not reported.

Results

Patient characteristics
Between January 1, 2006 and December 31, 2014, 69,935 patients with a diagnosis of empyema were identified (22,276 [32.8%] in California, 18 293 [26.1%] in New York, 22,284 [31.8%] in Florida, and 1,026 [10.1%] in Washington). Demographic and clinical characteristics are shown in Table 1a and 1b. Most patients were between the ages of 40-80 (70%), male (65%), and Caucasian (67%). All income quartiles were equally affected. Medicare patients represented 47.4% of our study population. Compared to patients from 2006, patients from 2014 were more likely to be older (p= < 0.0001). Patients presenting in 2014 were also more likely to have co-morbidities such as hypertension, obesity, peripheral vascular disease, pulmonary hypertension, and renal failure. However, the Charlson co-morbidity index was similar throughout the study period. The overall length of stay for each admission decreased over time.

30 & 90-day Readmission 
30 and 90-day readmission rates fluctuated throughout the study; however the overall trend was towards fewer readmissions in 2014 than 2006. Multivariate analysis of 30-day readmission Other race combines Hispanic, Asian or Pacific Islander, Native American, and Other listings in the State Inpatient Database.

showed that African Americans (OR 1.13 [1.052-1.21], p=< 0.0001), and Medicare/Medicaid recipients (OR 1.62 [1.46-1.81] and 1.69 [1.52-1.89]) were more likely to have a readmission. At 90 days there was no significant difference in readmission by age, sex, or race. Medicare and Medicaid recipients were still more likely to be readmitted (OR 1.82 [1.65-2.00] and 1.75 [1.59-1.93]). Comorbidities including anemia, congestive heart failure, chronic pulmonary disease, diabetes, liver disease, metastatic cancer, and renal failure were associated with increased risk of readmission at both 30 days and 90 days in multivariate analysis (Table 2a and 2b). Respiratory complications including pleural effusion, respiratory failure, postoperative infection, pneumonia, aspiration pneumonitis, and pneumothorax, and pulmonary abscess accounted for 24.5% of readmissions at 30 and 90 days.

Mortality
Mortality during the initial hospitalization was 11.3% in 2006 and 10.6% in 2014. This difference was not statistically significant. Multivariate analysis showed that age, African American race, congestive heart failure, liver disease, renal failure, peripheral vascular disease, Medicare/Medicaid payer, and fungal infection were associated with higher risk of death. When stratified by type of infection, mortality rates were lower in 2014 for Staphylococcus, Streptococcus, gram negative bacteria, but higher for fungal infections (Figure 2).

Treatment
Treatment strategies for this analysis included surgery, intercostal drainage, thoracentesis, and non-operative management. Since 2008 we noted a trend towards more surgery and intercostal drainage procedures and less thoracentesis and non-operative management. An additional change noted in 2011 was a trend towards more intercostal drainage and less surgical interventions. The overall rate of thoracentesis has slowly decreased (Figure 3). The mortality rates for surgery and intercostal drainage have been decreasing since 2006 while it has remained stable for non-operative management. 


Table 1a: Demographic Features of Patients with Empyema from the State Inpatient Database of California, Florida, New York and Washington, 2006-2014
 Abbreviations: N (Number of Patients); SNF (Short-Term Nursing Facility).
Other race combines Hispanic, Asian or Pacific Islander, Native American, and Other listings in the State Inpatient Database. 


Table 1b: Clinical Features of Patients with Empyema from the State Inpatient Database of California, Florida, New York and Washington, 2006-2014
Abbreviations: N (Number of Patients); CHF (Congestive Heart Failure); CPD (Chronic Pulmonary Disease); PVD (Peripheral Vascular Disease); PCD (Pulmonary Circulation Disorder).
Anemia combines chronic blood loss anemia and iron deficiency anemia. Diabetes combines complicated and uncomplicated diabetes.


Table 2a: Univariate and Multivariate Analysis of Factors Affecting 30 Day Readmission Rates in Patients with Empyema
*significant at p<0.01.
Other race combines Hispanic, Asian or Pacific Islander, Native American, and Other listings in the State Inpatient Database.


*significant at p<0.01
Other race combines Hispanic, Asian or Pacific Islander, Native American, and Other listings in the State Inpatient Database.
Table 2b: Univariate and Multivariate Analysis of Factors Affecting 90 Day Readmission Rates Among Patients with Empyema


Figure 2: Yearly Mortality Rates for Patients with Empyema by Infecting Organism


Figure 3: Yearly Trends in Treatment of Empyema

Discussion

In this retrospective cohort study of 69,935 inpatients with diagnosis of empyema in California, New York, Florida, and Washington from 2006 – 2014, most patients were between the ages of 40-80 (65%), Caucasian (67%), and had equal income quartile distribution. Most patients had a Charlson co-morbidity index between 1 and 5. Throughout our study period, the co-morbidity index followed a similar distribution, suggesting patients in 2014 were less likely to be sicker than those in 2006. Length of stay slowly decreased and was more likely to be less than 10 days in 2014 (p < 0.0001). This was accompanied by a trend towards more discharges to short term rehab or skilled nursing facilities in 2014 versus 2006 (38 vs 34%, p < 0.001). A similar trend has been explored recently by Barnett et al. [6] in Medicare patients. They found that between 2004 and 2011 hospital length of stay decreased from 6.3 to 5.7 days, while average length of stay at post-acute care facilities increased from 4.8 to 6.0 days. This trend is likely a result of the implementation of Medicare’s inpatient prospective payment system [6]. This national trend among Medicare patients has without question had an impact on our observations regarding length of stay.

The distribution of isolated pathogen changed during the time period of this analysis with less infections caused by staphylococcus aureus in 2014 (11% vs 7%, p < 0.0001). This is a new trend compared to data from 1996 – 2008 which showed that staphylococcus-related empyema had the largest absolute increase across age groups [7]. This may be attributed to increased MRSA (methicillin resistant Staphylococcus aureus) screening and contact precautions throughout hospitals.

Both 30 and 90-day readmission rates were statistically lower in 2014 compared to 2006 (22 vs 16% and 32 vs 21% respectively, p < 0.0001). This may be related to improved treatment during the initial hospitalization and higher number of discharges to skilled nursing facilities.

The Multicenter Intrapleural Sepsis Trials 1 and 2 (MIST1 and MIST2) analyzed the paradigm of empyema treatment. MIST1 showed that one could increase pleural fluid drainage with use of streptokinase but not reduce mortality [8]. MIST2 demonstrated that the combination of intrapleural TPA and DNAse had a statistically significant improvement in pleural drainage and a reduction in hospital stay. In addition, the need for surgical intervention at 3 months was reduced by 75% [9]. While the American Association of Thoracic Surgeons recommends against the routine use of fibrinolytics for complicated effusions and empyemas, there is a general trend throughout the pulmonary community to use them as a first line prior to surgery.3 In children for example, a study looking at national trend of empyema management noted an increase in the use of tube thoracostomy with fibrinolytics from 39% in 2009 to 53% in 2014. They also noted that the frequency of VATS peaked in 2009 (50.8%) and then dropped to 36.4% in 2014 [10]. In this analysis, we noted a similar trend. Surgical management of empyema reached a peak in 2009 (44%) and has been slowly down trending (39% in 2014). At the same time, intercostal drainage has been increasing since 2009 from 25% to 30% in 2014.

Our analysis found mortality rates for empyema to be lower in patients receiving surgical intervention or intercostal drainage with equal or greater mortality rates observed in those receiving thoracentesis and non-surgical management. An explanation for these findings could be that thoracentesis and non-surgical interventions are being reserved for an overall sicker population while intercostal drainage is becoming more effective with the more liberal use of TPA and DNAse. Despite some advances in the mortality rate overall, we did note a statistically higher mortality rate among African Americans and patients covered by Medicaid or Medicare.

Fungal empyemas carried the highest morbidity with a 16- 20% rate of death during admission. Streptoccocal infection was associated with the lowest mortality 3.8-7.2%. Staphylococcus was higher between 7.0 and 8.6%, and gram-negative bacteria at 8.5 – 12.8%.

There are several limitations to this study including those associated with large administrative databases. These limitations include sampling and coding errors and information limited to conditions and treatments defined by ICD-9-CM codes. Additionally, the infecting microbes defined by ICD-9-CM codes were unknown or not specified in about 65% of cases. It was not possible to differentiate between culture-negative empyema and empyema for which the causative organism was simply not recorded. However, these limitations would apply to all patients and thus should not negate the overall findings of this study. Furthermore, risk factors such as COPD, congestive heart failure, anemia, etc. have a spectrum of severity, which is not reflected by the coding within the database. Additionally, this database does not allow for the tracking of mortality for patients when they leave the hospital; therefore, we are only able to determine in-hospital mortality. Finally, the most recent data included in this study were 4 years old; therefore, the present study findings may not be reflective of the most recent characteristics and outcomes associated with empyema.

Conclusion

The use of less invasive approaches to treat empyema is increasing. This change has not been associated with any changes in hospital mortality; however, a trend towards decreased length of stay as well as 30- and 90-day readmissions is noted.

Acknowledgement

We would like to acknowledge Nishay Bhatnagar for his contribution to this paper in the form of a literature search.

References

  1. Colice GL, Curtis A, Deslauriers J, Heffner J, Light R, et al. Medicaland surgical treatment of parapneumonic effusions: an evidencebased guideline. Chest. 2000;118(4):1158-1171.
  2. Davies HE, Davies RJ, Davies CW, Group BTSPDG. Managementof pleural infection in adults: British Thoracic Society PleuralDisease Guideline 2010. Thorax. 2010 Aug;65 Suppl 2:ii41-53.
  3. Shen KR, Bribriesco A, Crabtree T, Denlinger C, Eby J, et al. TheAmerican Association for Thoracic Surgery consensus guidelines for the management of empyema. J Thorac Cardiovasc Surg. 2017Feb;153(6):e129-e146.
  4. Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, et al. Codingalgorithms for defining comorbidities in ICD-9-CM and ICD-10administrative data. Med Care. 2005 Nov;43(11):1130-1139.
  5. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidityindex for use with ICD-9-CM administrative databases. J ClinEpidemiol. 1992 Jun;45(6):613-619.
  6. Barnett ML, Grabowski DC, Mehrotra A. Home-to-Home Time -Measuring What Matters to Patients and Payers. N Engl J Med.2017 Jul;377(1):4-6.
  7. Grijalva CG, Zhu Y, Nuorti JP, Griffin MR. Emergence ofparapneumonic empyema in the USA. Thorax. 2011Aug;66(8):663-668.
  8. Maskell NA, Davies CW, Nunn AJ, Hedley EL, Gleeson FV, et al. U.K.Controlled trial of intrapleural streptokinase for pleural infection.N Engl J Med. 2005 Mar;352(9):865-874.
  9. Rahman NM, Maskell NA, West A, Teoh R, Arnold A, et al.Intrapleural use of tissue plasminogen activator and DNase inpleural infection. N Engl J Med. 2011 Aug;365(6):518-526.
  10. Dorman RM, Vali K, Rothstein DH. Trends in treatment ofinfectious parapneumonic effusions in U.S. children’s hospitals,2004-2014. J Pediatr Surg. 2016 Jun;51(6):885-890.


Copyright © 2020 Boffin Access Limited.